When I first got interested in Stirling engines one of the things that surprised me was the number of variables. For such a simple seeming engine, there is a whole lot going on. My strategy then was to control as many variables as possible, something I believe my epoxy LTD displacer chamber does well.
Thermal bridging is substantially reduced by the epoxy construction, the XPS foam displacer is about the best practical insulation that can be managed, and most importantly, it's as close to an on/off switch as I've seen. When the displacer is at rest on either hot or cold plate it is effectively "shut off". It subsequently achieves nearly the same results as PV=nRT in testing.
There is one major hurdle, the timing/phasing of displacer to power piston. The NASA and other free piston Stirling engines are surely tuned to operate at an ideal resonance for best performance, but this is partly what leads to their incredible complexity and expense.
So how do we control this variable? Drive the displacer, and let the piston be free to respond. The rpm of the displacer, work load, and the weight of the piston can then be adjusted to peak power/efficiency for the performance characteristics of the displacer chamber (a fixed variable). Or for fully crank based engines the observed piston response time can be used to set displacer advance.
This video is showing just that, the piston is only responding to the displacer and the phasing is naturally ideal, without any difficulty at all.
https://youtube.com/shorts/AvZ_GvFu8xE? ... i7idUbQLqj
Free the piston, but not the displacer!
-
matt brown
- Posts: 752
- Joined: Thu Feb 10, 2022 11:25 pm
Re: Free the piston, but not the displacer!
The Donald Rumsfeld quote comes to mind.
Indeed, the idea is to 'control' as many variables as possible, but this is best achieved thru a good design which ovoids conflicting issues from the get-go. A good example is isothermal heating which doesn't have to be beat to death since the obvious flaw is simply that the time constraint of heating limits output (another time function).
The order of merit that NASA employs in their parametric studies is unique and rarely includes any DIY reality check near the top (cost, complexity, etc). I usually lump any NASA type crap amongst the endless crap in PTO history.VincentG wrote: ↑Sat Jun 15, 2024 11:34 am There is one major hurdle, the timing/phasing of displacer to power piston. The NASA and other free piston Stirling engines are surely tuned to operate at an ideal resonance for best performance, but this is partly what leads to their incredible complexity and expense.
Beale gave us the FP displacer in the 1960s and Martini gave us the FP piston in the 1980s. Similar driving vs driven issues, most FP schemes suffer from a lead vs lag issue that is only valid across 1/2 the cycle (later Martini schemes included a small driven piston).VincentG wrote: ↑Sat Jun 15, 2024 11:34 am So how do we control this variable? Drive the displacer, and let the piston be free to respond. The rpm of the displacer, work load, and the weight of the piston can then be adjusted to peak power/efficiency for the performance characteristics of the displacer chamber (a fixed variable). Or for fully crank based engines the observed piston response time can be used to set displacer advance.
This video is showing just that, the piston is only responding to the displacer and the phasing is naturally ideal, without any difficulty at all.
https://youtube.com/shorts/AvZ_GvFu8xE? ... i7idUbQLqj
Re: Free the piston, but not the displacer!
Matt can you expand on the 1/2 cycle issue?Similar driving vs driven issues, most FP schemes suffer from a lead vs lag issue that is only valid across 1/2 the cycle (later Martini schemes included a small driven piston).
-
matt brown
- Posts: 752
- Joined: Thu Feb 10, 2022 11:25 pm
Re: Free the piston, but not the displacer!
AFAIK this was the last Martini version
Note the small conduit from bounce space to piston 'region'. The whole Martini spin is from a DOE grant (via NASA) and akin much BS from similar origins (grantmeisters). This evolved in early 1980s (during early electronics rage) and took advantage of lame DOE (these guys can't tie their shoes). Assume this scheme is to produce steady electricity and let's forget about a variable output. Thus, lacking modern electronics, how do you oscillate the linear generator at a constant output from a variable input force ??? If we assume that the generator is oscillating at a fixed speed then the gas is also expanding and compressing at a fixed speed. Appears sweet, but a nasty rears its ugly head that a unicorn ignores...as the gas expands, the force driving this expansion decreases as the volume increases, whereby the work output from gas expansion varies thruout each stroke. Any clever scheme with springs and bounce gas or whatever will be hard taxed to maintain uniform force to generator during each variable force from gas expansion.
The basic scheme approximates...1/2 the expansion gas force drives the generator while 1/2 the expansion gas force is retained for the 'return trip' via bounce gas, springs, etc. Indeed, an extremely crude reduction of the basic scheme, but it should convey the basic idea and myriad of issues. Here's the current FP scheme from wiki
This is only possible due to modern electronics without which there would be no output, nor cycle, and pondering how the force is equalized between strokes could keep you up at night. In the DIY world, we never have to worry about this stupid BS thanks to the magic of the flywheel...
- Martini.jpg (9.52 KiB) Viewed 1865 times
Note the small conduit from bounce space to piston 'region'. The whole Martini spin is from a DOE grant (via NASA) and akin much BS from similar origins (grantmeisters). This evolved in early 1980s (during early electronics rage) and took advantage of lame DOE (these guys can't tie their shoes). Assume this scheme is to produce steady electricity and let's forget about a variable output. Thus, lacking modern electronics, how do you oscillate the linear generator at a constant output from a variable input force ??? If we assume that the generator is oscillating at a fixed speed then the gas is also expanding and compressing at a fixed speed. Appears sweet, but a nasty rears its ugly head that a unicorn ignores...as the gas expands, the force driving this expansion decreases as the volume increases, whereby the work output from gas expansion varies thruout each stroke. Any clever scheme with springs and bounce gas or whatever will be hard taxed to maintain uniform force to generator during each variable force from gas expansion.
The basic scheme approximates...1/2 the expansion gas force drives the generator while 1/2 the expansion gas force is retained for the 'return trip' via bounce gas, springs, etc. Indeed, an extremely crude reduction of the basic scheme, but it should convey the basic idea and myriad of issues. Here's the current FP scheme from wiki
- FP wiki.jpg (22.57 KiB) Viewed 1865 times
This is only possible due to modern electronics without which there would be no output, nor cycle, and pondering how the force is equalized between strokes could keep you up at night. In the DIY world, we never have to worry about this stupid BS thanks to the magic of the flywheel...
Re: Free the piston, but not the displacer!
Matt is that Martini scheme driven displacer/free piston?
Doesn't this apply to any piston engine?Appears sweet, but a nasty rears its ugly head that a unicorn ignores...as the gas expands, the force driving this expansion decreases as the volume increases, whereby the work output from gas expansion varies thruout each stroke.
-
matt brown
- Posts: 752
- Joined: Thu Feb 10, 2022 11:25 pm
Re: Free the piston, but not the displacer!
yes
Doesn't this apply to any piston engine?Appears sweet, but a nasty rears its ugly head that a unicorn ignores...as the gas expands, the force driving this expansion decreases as the volume increases, whereby the work output from gas expansion varies thruout each stroke.
[/quote]
Yes, and things would be very limited without the magical flywheel. Newton died in 1726 and I've often wondered how he would address a weak force (piston) adding to a strong force (flywheel) during expansion vs the obvious opposite during compression.
Re: Free the piston, but not the displacer!
VincentG, I think it would be completely acceptable to drive a displacer with a small efficient electric motor, or solenoid, have a "free piston" linear generator to get electricity and use some of it to power the displacer. That is what the flywheel crank mechanism does anyway.
The displacer could be driven at different speeds, and strokes to vary the power out.
It could produce delay at each end, and shorter transit times.
It probably will be less efficient than a crank system, the above benefits hopefully out weighing this.
Many modifications could be done to just software.
My thoughts seem to add springs to retain harmonic motion, which would tend to limit it to a designed speed.
It could be made to easily exchange springs.
Yes great. Probably learn in the process.
The displacer could be driven at different speeds, and strokes to vary the power out.
It could produce delay at each end, and shorter transit times.
It probably will be less efficient than a crank system, the above benefits hopefully out weighing this.
Many modifications could be done to just software.
My thoughts seem to add springs to retain harmonic motion, which would tend to limit it to a designed speed.
It could be made to easily exchange springs.
Yes great. Probably learn in the process.
Re: Free the piston, but not the displacer!
Is your Dremel drill driving the displacer variable speed? You may have mentioned that but if so I've forgotten.
Anyway, if, just for kicks, we were to apply the 90° phase advance angle theory to this problem, we would assume that the power piston with a given fixed mass would have a "preferred" frequency of oscillation and that ideally the displacer should...
Well, not "should" perhaps. I'll try putting this another way.
IF the displacer frequency could be adjusted up and down so as to locate and match the preferred "natural frequency" of the power piston there should be a quite noticeable response once the two are "in resonance".
Unfortunately, I suppose that would not be the outcome you were probably trying to achieve, some control over the piston, but that might be possible by pausing the displacer briefly and start it up again to put it's motion out of phase with the power piston or in phase, or to one degree or another move it away from the resonant frequency.
Anyway, if the Dremel drill had infinitely variable speed it might be quite easy to "tune in" to the "natural frequency" of the power piston, whatever that might be, by simply adjusting the speed of the Dremel tool until you hit a speed the piston responds to in some presumably dramatic way, akin to Tesla's "earthquake machine" I suppose.
Anyway, if, just for kicks, we were to apply the 90° phase advance angle theory to this problem, we would assume that the power piston with a given fixed mass would have a "preferred" frequency of oscillation and that ideally the displacer should...
Well, not "should" perhaps. I'll try putting this another way.
IF the displacer frequency could be adjusted up and down so as to locate and match the preferred "natural frequency" of the power piston there should be a quite noticeable response once the two are "in resonance".
Unfortunately, I suppose that would not be the outcome you were probably trying to achieve, some control over the piston, but that might be possible by pausing the displacer briefly and start it up again to put it's motion out of phase with the power piston or in phase, or to one degree or another move it away from the resonant frequency.
Anyway, if the Dremel drill had infinitely variable speed it might be quite easy to "tune in" to the "natural frequency" of the power piston, whatever that might be, by simply adjusting the speed of the Dremel tool until you hit a speed the piston responds to in some presumably dramatic way, akin to Tesla's "earthquake machine" I suppose.
Re: Free the piston, but not the displacer!
It sounds logical to be able to adjust the frequency of the driver to the frequency of the mass spring system.
Re: Free the piston, but not the displacer!
That is the whole point if this, to bring the cycle into a natural resonance with a weighted piston eventually.Unfortunately, I suppose that would not be the outcome you were probably trying to achieve, some control over the piston, but that might be possible by pausing the displacer briefly and start it up again to put it's motion out of phase with the power piston or in phase, or to one degree or another move it away from the resonant frequency.
The weighted piston will allow the cycle to extend into adiabatic expansion/compression.
Re: Free the piston, but not the displacer!
Remember, it's only adiabatic if the temperatures surrounding the gas are the same. Even very fast processes will transfer some heat if there is a temperature difference, the bigger difference the more transfers. Adiabatic processes change the temperature of the gas, thus effecting a difference.
Re: Free the piston, but not the displacer!
You had a comment recently that mentioned "quasi-adiabatic" and "quasi-isothermal". Maybe I should always use those terms as it's always what I mean.
As far as the difference, I guess it's the tipping point between the two, though I'll always aim for 75% of ideal. Otherwise it's easy to pretend you are getting the desired results even if it's not so.
As far as the difference, I guess it's the tipping point between the two, though I'll always aim for 75% of ideal. Otherwise it's easy to pretend you are getting the desired results even if it's not so.
Re: Free the piston, but not the displacer!
Hey Matt where did that Martini design end up?
Re: Free the piston, but not the displacer!
I like quasi isothermal. That implies that it's important to transfer heat to get functionality. Quasi adiabatic just seems to to be an oxymoron, or not. They were both coined to be identical to highlight the compromise all processes will have.
I guess the choice of which to use would be determined by which effect dominates. Quasi adiabatic would be used for diesel ignition and a fire piston. Quasi isothermal for slower processes that need heat transfer to work, such as Stirling Engine expansion and compression.
Just an attempt to be less misleading.
Adiabatic bounce would be a middle ground as the bounce quickly decays, probably from heat transfer, loss and hysteresis, on each bounce. And leakage and friction. I will still call it adiabatic bounce, because the bounce could ideally be adiabatic even though it isn't. Like calling it a Stirling Engine and a Stirling Cycle. Ideal or not. A real engine or a mathematical model and goal to head towards.
The indicator diagram for Stirling Engines, as they run faster and faster seem to shrink down closer closer to a single zero area, a zero power line. I suppose it is an adiabatic line. Speed and heat flow must be compromised to get to an engines desired operating level. It will have a big effect on power produced, and efficiency. Fastest isn't bestest. Slowest isn't bestest.
All operations are both quasi, isothermal and adiabatic, at any speed. Even fire pistons and diesel engines, that is why diesel engines get hot when running. A fire piston would get hot too if used repeatedly. Like a bicycle pump.
I guess the choice of which to use would be determined by which effect dominates. Quasi adiabatic would be used for diesel ignition and a fire piston. Quasi isothermal for slower processes that need heat transfer to work, such as Stirling Engine expansion and compression.
Just an attempt to be less misleading.
Adiabatic bounce would be a middle ground as the bounce quickly decays, probably from heat transfer, loss and hysteresis, on each bounce. And leakage and friction. I will still call it adiabatic bounce, because the bounce could ideally be adiabatic even though it isn't. Like calling it a Stirling Engine and a Stirling Cycle. Ideal or not. A real engine or a mathematical model and goal to head towards.
The indicator diagram for Stirling Engines, as they run faster and faster seem to shrink down closer closer to a single zero area, a zero power line. I suppose it is an adiabatic line. Speed and heat flow must be compromised to get to an engines desired operating level. It will have a big effect on power produced, and efficiency. Fastest isn't bestest. Slowest isn't bestest.
All operations are both quasi, isothermal and adiabatic, at any speed. Even fire pistons and diesel engines, that is why diesel engines get hot when running. A fire piston would get hot too if used repeatedly. Like a bicycle pump.